Thinking, arguing, and counter-arguing: the effects of issue relevance and minority influence on the choice of argumentation strategies

The relationship between argumentation and reasoning has received much attention in educational psychology, as argumentation is considered the tool that supports any cognitive activity. Nevertheless, being confronted with conflicting points of view does not guarantee a cognitive progress, since different argumentation strategies can reflect more or less accurate cognitive elaboration. In that, we make a parallel between argumentation and hypothesis testing tasks, where confirmatory biases, anchoring effects, etc., are examples of a general focussing effect (i.e., a tendency to consider only the elements already present in one’s cognitive field and to ignore any divergent information), while falsification of alternatives witnesses a decentering tendency that ensures the use/acquisition of more advanced cognitive skills. We propose that the same focussing and decentering tendencies can be retrieved also in the field of argumentation. In fact, when arguing about an issue, speakers can hold their position either a) enforcing some statements they already agree with, or b) trying to confute alternative standpoints. Developing this parallel, we also predict that socio-cognitive conditions that are proved to foster decentering in hypothesis testing, will favour the appearance of counter-arguing strategies. In particular, we hypothesize that high personal relevance, and confrontation with a minority source of influence, should enhance the use of counter-arguments, whereas low relevance of the issue, and exposure to majority influence, should favour the enforcement of already shared arguments. In a 2 (relevance: high vs. low) x 2 (source of influence: majority vs. minority) experimental design, 114 university students were asked to develop three arguments in order to express their point of view about the introduction of a final comprehensive exam, choosing between confirmatory and disconfirmatory arguments. Results confirm the predicted source*relevance interaction. Implication for the design of argumentation activities in education will be discussed.